Key takeaways: The main image wins the click; slots two through seven win the sale. Give every image one job: in use, scale, key feature, what is included, the objection from your reviews, and the comparison. Follow the main-image rules exactly, build for a phone screen, and test the main image with Manage Your Experiments before you trust it.
Why images decide more than copy
On a search results page the shopper sees a thumbnail, a price, a star rating and a truncated title. The thumbnail is the only thing that is fully yours to control at that moment. On the detail page, most shoppers on a phone swipe through the gallery before they read a single bullet. The image set is the sales conversation, and the copy is the footnote.
That is why image work sits at the front of the fix list for two of the most common problems on Amazon: low click-through rate, which is mostly a main image and price problem, and low conversion rate, which is mostly a gallery and reviews problem. The How to Improve Amazon Click-Through Rate (CTR) article covers the first; this guide covers both.
How a gallery gets built
Read this flowchart as text
- Brief: product, buyer, top 5 objections from reviews
- Main image: white background, product fills the frame
- Slots 2 to 7: one job each use, scale, feature, box, objection, compare
- Meets Amazon image requirements and policy?
- Upload, check order and mobile rendering
- CTR and unit session percentage after 2 to 4 weeks?
- Keep, apply the pattern to the rest of the range
- A/B test the main image with Manage Your Experiments
The requirements that apply to every image
Amazon publishes technical requirements and enforces them at upload and in later audits. The ones that matter most:
- Format. JPEG is the safe choice. PNG, GIF and TIFF are accepted, but JPEG avoids most rendering issues.
- Size. At least 1,000 pixels on the longest side so the zoom function works. Larger is better; 2,000 pixels or more gives a clean zoom, and Amazon accepts images up to 10,000 pixels.
- Aspect. Square or near-square images display most consistently across desktop, mobile and the app.
- Color. sRGB or CMYK; sRGB is what displays correctly.
- Content. The product must be what is sold, in full, and must not be shown with accessories that are not included unless they are clearly labeled. No nudity, no offensive material, no Amazon logos or badges, no text that imitates Amazon's interface.
Slot-specific rules are stricter for the main image and looser for the rest, covered below.
Slot 1: The main image
The main image has the strictest rules and the biggest effect. It must be on a pure white background, RGB 255, 255, 255. It must show only the product being sold, filling most of the frame, with Amazon's guidance suggesting the product occupies around 85% of the image area. No text, no graphics, no logos other than what is printed on the product, no props, no packaging unless the packaging is the product, no multiple angles combined into one frame.
Within those rules there is a surprising amount of room:
- Angle. A slight three-quarter view usually reads better than dead-on flat, because it shows depth. Test it.
- Fill. A product that fills the frame edge to edge is bigger in the thumbnail than one floating in white space. The 85% guidance is a minimum, not a target.
- Packaging. For some categories, especially consumables and multipacks, showing the retail package alongside the product is allowed and can lift click-through because it signals quantity. Check your category's style guide.
- Quantity. A multipack should look like a multipack in the thumbnail. A single unit that is actually a pack of six loses clicks to the competitor whose thumbnail shows all six.
- Color and contrast. On a results page full of white backgrounds, the product's own color and shadow are what separate it from the tile next to it.
The main image is the one to A/B test first, because it acts on click-through rate and click-through rate acts on rank. A small lift in CTR on a high-traffic term compounds for months.
Slots 2 to 7: One job each
Amazon shows a limited number of gallery thumbnails on the detail page, and if you add a video it takes one of those positions. Plan for seven images and assume the seventh may be pushed behind a "see more" control on some devices. Every slot needs a job, and the jobs should follow the order of the shopper's questions.
Slot 2: The product in use. The lifestyle image. Someone using the product in the setting the buyer imagines. Its job is to answer "is this for me?" in a second. Real environments, real hands, correct scale, no stock photo feel. The Amazon Lifestyle Images: Making the Product Real article covers what makes these work.
Slot 3: Scale and dimensions. The most common reason for returns is "it was smaller than I expected" or "it did not fit". Show the product next to a familiar object or a person, with dimensions marked clearly. For anything that has to fit a space or another product, this slot is not optional.
Slot 4: The key feature or how it works. The infographic. One feature, one image, with a short label. If the product has three important features, that is three images, not one image with nine callouts. The Amazon Infographic Images: What to Show and What to Skip article covers what to show and what to leave out.
Slot 5: What is in the box. Lay out every item the buyer receives. This prevents the "I thought it came with..." review and reassures the shopper who is comparing bundles.
Slot 6: The objection. Read your one-star and two-star reviews and your competitors'. Whatever shoppers worry about most, this slot addresses directly: durability shown by a materials close-up, safety shown by a certification mark you actually hold, noise shown by a decibel figure you have measured, compatibility shown by a list of models it fits.
Slot 7: Comparison or range. Your own range side by side, or this product against the category's typical alternative on attributes you win. Never name or show a competitor's brand. Comparison against "standard" or "ordinary" versions is the safe framing.
Video, if you have it, usually earns a slot. Anything that moves, assembles, pours or transforms is better shown than described. The Amazon Product Video: Where It Appears and What Works article covers where video appears and what works.
Designing for the phone
Most shoppers see your gallery on a screen a few inches wide, swiping. This changes how images should be built:
- Text size. Any label or callout must be readable at thumbnail scale. If you have to squint on a phone, the text is decoration, not information. Three to six words per label, large.
- One focal point. Each image should be understood in under two seconds. Busy compositions are skipped.
- Consistent framing. A gallery where the product changes size and position from slot to slot feels cheap. Keep the product roughly the same scale and position across the infographic slots.
- Consistent style. Same background treatment, same color palette, same typography across all seven images, your A+ Content and your Brand Store. Consistency reads as quality.
- Test on a device. Open the live listing on a phone before you sign the images off. Not a preview: the live page.
Briefing a designer or photographer
Good images start with a good brief. The brief should contain:
- The product, its exact dimensions, and how it ships.
- Who buys it and what they use it for, in a sentence.
- The top five objections from reviews, ranked.
- The competitor listings you are up against, with notes on what their galleries do well and badly.
- The slot plan: seven slots, one job each, written out.
- Brand assets: logo, palette, fonts, any existing photography.
- What is not allowed: policy limits, claims you cannot make, badges you do not hold.
A designer with that brief produces images that sell. A designer with "make it look premium" produces images that look premium and do not answer any question.
Photography versus rendering
Both work, and most strong galleries use both. Photography wins for lifestyle, texture and anything where authenticity matters, such as food or fabric. Rendering wins for consistent product angles, exploded views, cutaways and products that are hard to light. Whichever you use, the main image must be a true representation of the product; a render that flatters the product into something the buyer does not receive earns returns and reviews that undo the gain.
Testing images on live traffic
Manage Your Experiments, available to Brand Registry sellers, lets you run two versions of the main image against each other on live traffic and reports which version converted better with a confidence figure. Use it. Test the main image first, because it acts on both click-through and conversion. Then test the second slot.
Tests need traffic. Amazon indicates eligibility per ASIN based on recent traffic, and low-volume products may not qualify. For those, run a before-and-after comparison on unit session percentage across four weeks each side, with the price held constant.
Change one thing per test. A new main image with a new angle, a new fill and new packaging tells you the set won or lost, not which element did it. The Manage Your Experiments: What to A/B Test First on Amazon article covers how to set tests up.
Common mistakes
- A main image that breaks the rules. Text, a badge, a prop, an off-white background. Amazon suppresses the listing, sometimes weeks after upload.
- Nine callouts on one infographic. Unreadable on a phone. One feature per image.
- No scale image. The leading cause of "smaller than expected" returns.
- Lifestyle images that are obviously stock. Shoppers notice, and it reads as a brand with nothing real to show.
- Ignoring the reviews. The objection slot is the one that changes conversion most, and it is the one most galleries skip.
- Inconsistent style across the page. Gallery, A+ Content and Store in three different looks.
- Never testing. An image set is a hypothesis until conversion data says otherwise.
- Changing images during a launch or an event. Do it in a quiet period so the effect can be read.
How WeSpark does it
Our creative work starts with a brief built from your reviews and your competitors' galleries, assigns one job to each of the seven slots, and measures the result on click-through and conversion. The processes below are the ones we follow.
Main images are A/B tested where the ASIN qualifies, and every image is checked on a live phone page before it is signed off. The gallery is designed together with A+ Content and the Brand Store so the page reads as one brand.
Find your next growth move.
An account audit, a 90-day roadmap, and a written scope before you pay.
Frequently asked questions
How many images should an Amazon listing have?
Fill all the slots Amazon shows in the gallery, which for most categories means seven, with video taking one of them if you have it. Fewer than five leaves questions unanswered. More than the visible set is fine but the extra images are seen by few shoppers, so put the important ones first.
What are the Amazon main image requirements?
Pure white background, the product only, filling most of the frame, no text, graphics, props or badges, and at least 1,000 pixels on the longest side so zoom works. Category style guides add rules on packaging and how the product may be shown. Breaking these gets the listing suppressed.
Which image should I test first?
The main image. It acts on click-through rate in search results and on conversion on the detail page, and click-through feeds rank. Test angle, fill or packaging one at a time using Manage Your Experiments if the ASIN qualifies.
Should I use lifestyle photos or infographics?
Both, in different slots. Lifestyle images answer "is this for me?" and belong in slot two. Infographics answer "how does it work and what do I get?" and belong in the middle slots. A gallery of only one type leaves half the shopper's questions open.
Can I put text on Amazon product images?
Not on the main image. On the secondary images, yes, and short labels are useful, but keep them large enough to read on a phone and keep claims within Amazon's policy: no unsupported superlatives, no pricing, no Amazon badges or logos.
How long before new images show a result?
Click-through changes show within one to two weeks on a product with steady traffic. Conversion changes need three to four weeks. Judge on unit session percentage and CTR from the Search Query Performance report, with price held constant.
